fedi-index: Kodning
Hvilke modeller skriver bedst kode: Arena-stemmer for kodning og webudvikling plus uafhængige SWE-bench- og Terminal-Bench-kørsler.
| # | Tier | Model | fedi-index | Percentil | Folk foretrækker | Løser opgaver | Værdi | Kilder | Opdeling af scoren |
|---|---|---|---|---|---|---|---|---|---|
| 15 | S | 1.462 ±2 | 95 | 1.477 | 1.423 | +102 | AT | › | |
| 16 | A | 1.461 ±5 | 95 | 1.463 | 1.457 | +52 | AT | › | |
| 20 | A | 1.454 ±2 | 93 | 1.469 | — | — | A | › | |
| 23 | A | 1.452 ±5 | 92 | 1.472 | 1.404 | +101◆ | AT | › | |
| 24 | A | 1.452 ±3 | 92 | 1.461 | 1.428 | +65 | AT | › | |
| 25 | A | 1.451 ±3 | 92 | 1.463 | 1.421 | +124◆ | AT | › | |
| 31 | A | 1.449 ±3 | 90 | 1.468 | 1.398 | +39 | AT | › | |
| 33 | A | 1.448 ±2 | 89 | 1.470 | 1.392 | — | AT | › | |
| 34 | A | 1.447 ±4 | 89 | 1.462 | — | +108◆ | A | › | |
| 37 | A | 1.444 ±3 | 88 | 1.454 | 1.396 | +49 | AE | › | |
| 38 | A | 1.443 ±2 | 87 | 1.457 | — | — | A | › | |
| 40 | A | 1.440 ±3 | 87 | 1.461 | 1.387 | +105◆ | AT | › | |
| 47 | A | 1.435 ±3 | 84 | 1.449 | — | +64 | A | › | |
| 50 | A | 1.434 ±3 | 83 | 1.443 | 1.391 | +35 | AE | › | |
| 53 | A | 1.432 ±3 | 82 | 1.441 | 1.387 | +46 | AE | › | |
| 54 | A | 1.432 ±4 | 82 | 1.446 | — | +35 | A | › | |
| 59 | A | 1.429 ±3 | 80 | 1.440 | 1.377 | +38 | AE | › | |
| 60 | B | 1.429 ±4 | 80 | 1.442 | — | +45 | A | › | |
| 61 | B | 1.428 ±2 | 79 | 1.442 | — | +74 | A | › | |
| 67 | B | 1.424 ±3 | 77 | 1.438 | — | +88 | A | › | |
| 74 | B | 1.419 ±5 | 75 | 1.428 | 1.376 | +41 | AE | › | |
| 75 | B | 1.419 ±4 | 74 | 1.432 | — | +80 | A | › | |
| 82 | B | 1.416 ±2 | 72 | 1.428 | — | +15 | A | › | |
| 84 | B | 1.413 ±2 | 71 | 1.429 | 1.373 | +58 | AT | › | |
| 87 | B | 1.413 ±9 | 70 | 1.426 | — | +36 | A | › | |
| 90 | B | 1.411 ±4 | 69 | 1.420 | 1.368 | +26 | AE | › | |
| 91 | B | 1.411 ±3 | 69 | 1.423 | — | +76 | A | › | |
| 92 | B | 1.411 ±10 | 69 | 1.431 | — | +30 | A | › | |
| 93 | B | 1.409 ±7 | 68 | 1.423 | — | +42 | A | › | |
| 95 | B | 1.407 ±2 | 68 | 1.419 | — | +28 | A | › | |
| 98 | B | 1.404 ±2 | 67 | 1.416 | — | +33 | A | › | |
| 100 | B | 1.404 ±5 | 66 | 1.416 | — | +27 | A | › | |
| 105 | B | 1.400 ±3 | 64 | 1.412 | — | +50 | A | › | |
| 106 | B | 1.400 ±2 | 64 | 1.412 | — | +36 | A | › | |
| 112 | B | 1.396 ±5 | 62 | 1.408 | — | +76 | A | › | |
| 113 | B | 1.395 ±3 | 61 | 1.407 | — | +64 | A | › | |
| 115 | B | 1.392 ±7 | 61 | 1.404 | — | +42 | A | › | |
| 117 | B | 1.390 ±3 | 60 | 1.402 | — | +31 | A | › | |
| 118 | B | 1.390 ±7 | 60 | 1.402 | — | +48 | A | › | |
| 119 | B | 1.390 ±6 | 59 | 1.402 | — | +67 | A | › | |
| 121 | B | 1.389 ±10 | 59 | 1.402 | — | +34 | A | › | |
| 122 | B | 1.389 ±4 | 58 | 1.400 | — | +65 | A | › | |
| 123 | B | 1.388 ±3 | 58 | 1.399 | — | +14 | A | › | |
| 125 | B | 1.387 ±7 | 57 | 1.399 | — | +22 | A | › | |
| 126 | B | 1.387 ±3 | 57 | 1.398 | — | +71◆ | A | › | |
| 127 | B | 1.385 ±4 | 57 | 1.396 | — | +26 | A | › | |
| 128 | B | 1.384 ±2 | 56 | 1.395 | — | — | A | › | |
| 129 | B | 1.384 ±3 | 56 | 1.395 | — | +17 | A | › | |
| 132 | B | 1.381 ±6 | 55 | 1.392 | — | +46 | A | › | |
| 133 | B | 1.380 ±9 | 54 | 1.392 | — | +2 | A | › | |
| 136 | B | 1.378 ±11 | 53 | 1.390 | — | -13 | A | › | |
| 138 | B | 1.377 ±11 | 53 | 1.389 | — | +15 | A | › | |
| 139 | B | 1.376 ±5 | 52 | 1.387 | — | +56 | A | › | |
| 140 | B | 1.373 ±7 | 52 | 1.384 | — | +33 | A | › | |
| 141 | B | 1.372 ±18 | 52 | 1.387 | — | +39 | A | › | |
| 143 | B | 1.372 ±10 | 51 | 1.383 | — | +30 | A | › | |
| 144 | B | 1.371 ±2 | 51 | 1.384 | — | -6 | A | › | |
| 146 | B | 1.370 ±4 | 50 | 1.383 | — | -20 | A | › | |
| 148 | C | 1.368 ±4 | 49 | 1.378 | — | -18 | A | › | |
| 153 | C | 1.364 ±4 | 48 | 1.374 | — | +30 | A | › | |
| 158 | C | 1.362 ±10 | 46 | 1.372 | — | +21 | A | › | |
| 162 | C | 1.358 ±10 | 44 | 1.369 | — | -15 | A | › | |
| 163 | C | 1.358 ±6 | 44 | 1.368 | — | +16 | A | › | |
| 166 | C | 1.354 ±6 | 43 | 1.364 | — | -24 | A | › | |
| 167 | C | 1.353 ±9 | 43 | 1.363 | — | -5 | A | › | |
| 169 | C | 1.350 ±4 | 42 | — | 1.372 | +43 | T | › | |
| 174 | C | 1.346 ±18 | 40 | 1.358 | — | -3 | A | › | |
| 175 | C | 1.346 ±9 | 40 | 1.355 | — | +16 | A | › | |
| 177 | C | 1.344 ±6 | 39 | 1.353 | — | +14 | A | › | |
| 178 | C | 1.343 ±8 | 39 | 1.352 | — | +20◆ | A | › | |
| 183 | C | 1.341 ±4 | 37 | 1.349 | — | -20 | A | › | |
| 186 | C | 1.335 ±16 | 36 | 1.346 | — | +4 | A | › | |
| 189 | C | 1.328 ±6 | 35 | 1.336 | — | -42 | A | › | |
| 190 | C | 1.324 ±5 | 35 | 1.334 | — | -6 | A | › | |
| 192 | C | 1.321 ±17 | 34 | 1.330 | — | -30 | A | › | |
| 195 | C | 1.316 ±14 | 33 | 1.324 | — | -56 | A | › | |
| 196 | C | 1.313 ±8 | 33 | 1.320 | — | — | A | › | |
| 203 | C | 1.286 ±10 | 30 | 1.291 | — | -14 | A | › | |
| 230 | D | 1.204 ±14 | 21 | 1.204 | — | -117 | A | › |
AArena (menneskelige stemmer)TTerminal-BenchEEpoch AI (benchmarks) ◆ Pareto-front: ingen er både bedre og billigere
fedi-index er en Elo-ækvivalent på en fast skala: sammenlignelig mellem måneder inden for én metodikversion. Niveau S–D: efter percentil i udsnittet og ikke bedre, end konfidensintervallet tillader (grænser i metodikken). »Foreløbig« — kun én kildefamilie har målt modellen her. Metodik
Indlejr på dit website
Tabellen som widget til dit website: den opdaterer sig selv og bevarer kildehenvisningerne.